36 BULLETIN 642, U. S. DEPARTMENT OF AGRICULTURE. | 
ations in the quantity of sediment as well as in the bacterial count, 
but on the whole there was a strikingly close relation. Plate V shows - 
that after the morning milking on October 22 there was a decided 
drop in the quantity of sediment as well as in the bacterial count. 
On that date the manure was removed from the stable in the after- 
noon. That night’s milk showed decidedly less sediment and a lower 
bacterial count, as did also both milkings of October 23. The results 
show that there is some direct relation between the sediment and 
bacterial count which always may be influenced by the variable num- 
ber of bacteria in the manure. It is therefore quite evident that the 
sediment test is of value only under certain conditions.. 
Bacterial counts shown in Plate V are particularly interesting 
since they show that a very large quantity of sediment, in other 
words, manure, is necessary to create high counts provided the princi- 
pal source of contamination is manure. In the experiment in which 
the manure was not removed nor the cows cleaned for nine days, the 
highest count obtained was 1,550,000, and the next highest was 
1,150,000 bacteria per cubic centimeter. The other counts ranged 
from 6,600 to 450,000 bacteria per cubic centimeter. After the 
manure was removed the experiment was repeated under similar con- 
ditions for 10 days. During the latter period the highest count ob- » 
tained was 82,000 and the lowest 6,600. In both experiments steril- 
ized utensils were used; consequently, the contamination was due 
principally to the introduction of manure into the milk. The figures 
indicate that counts above 200,000 per cubic centimeter in milk usu- 
ally should not be attributed entirely to contamination from manure. 
The number of bacteria in 57 samples of fresh cow manure was 
determined, and the results are shown in Table 16. The range in 
number of bacteria per gram was from 2,900,000 to 690,000,000, the 
average number per gram being 49,645,614. From these results it can © 
be assumed, therefore, that the average bacterial content of fresh 
manure is about 50,000,000 per gram. From that number it is possi- 
ble to calculate the number of bacteria added to each cubic centimeter 
of a pint of milk through contamination by definite quantities of 
manure. Table 17 shows the number of bacteria that would be added 
to each cubic centimeter of milk if fresh manure were added in quan- 
tities varying from 0.5 to 0.01 of a gram, assuming an average gram 
sample to contain 50,000,000 bacteria. The figures show that 0.5 
gram of fresh manure would add 52,854 bacteria to each cubic centi- 
meter of a pint of milk, while 0.01 of a gram of manure would add 
1,057. 
na 
